Académique Documents
Professionnel Documents
Culture Documents
SAFE
Integrated
Analysis and Design of Slabs
by
The Finite Element Method
VERIFICATION MANUAL
COMPUTERS &
STRUCTURES
INC.
The computer program SAFE and all associated documentation are pro-
prietary and copyrighted products. Worldwide rights of ownership rest
with Computers and Structures, Inc. Unlicensed use of the program or re-
production of the documentation in any form, without prior written
authorization from Computers and Structures, Inc., is explicitly prohib-
ited.
Further information and copies of this documentation may be obtained
from:
Introduction 1
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Organization of Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
i
SAFE Verification Manual
Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Modeling Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Comparison of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
File Reference. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
ii
Table of Contents
Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Modeling Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Comparison of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
File Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
iii
SAFE Verification Manual
References 125
iv
Introduction
Overview
SAFE is a reinforced concrete slab and basemat analysis and design program based
on the finite element method. This manual contains a collection of examples verify-
ing the accuracy and applicability of the program.
The first seven examples verify the accuracy of the elements and the solution algo-
rithms used in SAFE. These examples compare displacements and member internal
forces predicted by SAFE with known theoretical solutions for various slab support
and load conditions.
The next seven examples verify the applicability of SAFE in calculating design mo-
ments in slabs by comparing results for practical slab geometries with experimental
results and/or results using ACI 318-95 recommendations.
The last three examples verify the design algorithms used in SAFE for flexural and
shear design, using the ACI 318-95 recommendations, by comparing SAFE results
with hand calculations.
Organization of Manual
This manual includes one chapter for each example. A typical chapter consists of
the following sections:
Organization of Manual 1
SAFE Verification Manual
File Reference: This section contains referenced filenames associated with in-
dividual examples. These files are provided with the complete SAFE package
so that the user can independently cross reference and validate data reported in
this manual.
2 Organization of Manual
Example 1
Data
Plate size, a b = 360 in 240 in
Plate thickness, T = 8 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.3
Load Cases:
Data 3
SAFE Verification Manual
Modeling Procedure
To test convergence, the problem is analyzed employing three mesh sizes, 4 4 ,
8 8 , and 12 12 , as shown in Figure 1-2. The slab is modeled using thin plate ele-
ments in SAFE. The simply supported edges are modeled as line supports with a
large vertical stiffness. Three load cases are considered. Self-weight is not included
in these analyses.
To obtain design moments, the plate is divided into three strips two edge strips
and one middle strip each way, based on the ACI 318-95 definition of design
strip widths for a two-way slab system as shown in Figure 1-3. For comparison with
the theoretical results, load factors of unity are used and each load case is processed
as a separate load combination.
Comparison of Results
The deflections of four different points for three different mesh refinements are
shown for the three load cases in Table 1-1. The theoretical solutions based on Na-
viers formulations are also shown for comparison. This figure also shows the con-
vergence of the SAFE solution with mesh refinement. It can be observed from
Table 1-1 that the deflection obtained from SAFE converges monotonically to the
theoretical solution with mesh refinement. Moreover, the agreement is good for
even the coarse mesh (4 4). The numerically obtained local-moments and local-
shears at critical points are compared with that of the theoretical values in Table 1-2
and Table 1-3, respectively. Close agreement has been observed. A noticeable dis-
crepancy, however, occurs for the case of the point load (load case PL) in the region
close to the application of the point load, where the theoretical model has a singu-
larity. The average strip-moments for the load cases are compared with the theoreti-
cal average strip-moments in Table 1-4, where excellent agreement can be ob-
served. It should be noted that in calculating the theoretical solution, a sufficient
number of terms from the series are taken into account to achieve the accuracy of
the theoretical solutions.
File Reference
The files for the 4 4 , 8 8 , and 12 12 meshes are S01a.FDB, S01b.FDB, and
S01c.FDB, respectively. These files are included in the SAFE package.
4 Modeling Procedure
Example 1 Simply-Supported Rectangular Plate
Figure 1-1
Rectangular Plate Example 1
File Reference 5
SAFE Verification Manual
Figure 1-2
SAFE Meshes for Rectangular Plate
6 File Reference
Example 1 Simply-Supported Rectangular Plate
Edge Strip
Middle Strip
X-Strips
X
Y-Strips
Figure 1-3
SAFE Definition of Design Strips
File Reference 7
SAFE Verification Manual
Table 1-1
Comparison of Displacements
8 File Reference
Example 1 Simply-Supported Rectangular Plate
Load Moments
Case (kip-in/in)
Location
Mx My M xy
Table 1-2
Comparison of Local Moments
File Reference 9
SAFE Verification Manual
Vx Vy
Table 1-3
Comparison of Local Shears
10 File Reference
Example 1 Simply-Supported Rectangular Plate
y = 120
Middle 2.703 2.770 2.782 2.792
y = 120
Middle 3.521 3.371 3.358 3.307
y = 120
Middle 3.083 3.200 3.221 3.200
Table 1-4
Comparison of Average Strip Moments
File Reference 11
.
Example 2
Description
A fully clamped rectangular plate, as shown in Figure 2-1, is analyzed for three dif-
ferent load conditions. A theoretical solution to this problem, employing a single
series (Lvys solution) is given in Timoshenko and Woinowsky (1959). The nu-
merically computed deflections obtained from SAFE are compared with the theo-
retical values.
Data
Plate size, a b = 360 240
Plate thickness, T = 8 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.3
Data 13
SAFE Verification Manual
Load Cases:
Modeling Procedure
To test convergence, the problem is analyzed using three mesh sizes, 4 4 , 8 8 ,
and 12 12 , as shown in Figure 1-2. The plate is modeled using thin plate elements
available in SAFE. The fixed edges are modeled as line supports with large vertical
and rotational stiffnesses. The self-weight of the plate is not included in any of the
load cases.
Comparison of Results
The numerical displacements obtained from SAFE are compared with those ob-
tained from the theoretical solution in Table 2-1. The theoretical results are based
on tabular values given in Timoshenko and Woinowsky (1959). A comparison of
deflections for the three load cases shows a fast convergence to the theoretical val-
ues with successive mesh refinement.
File Reference
The files for the 4 4 , 8 8 , and 12 12 meshes are S02a.FDB, S02b.FDB, and
S02c.FDB, respectively. These files are included in the SAFE package.
14 Modeling Procedure
Example 2 Rectangular Plate with Built-In Edges
Figure 2-1
Rectangular Plate with All Edges Clamped
File Reference 15
SAFE Verification Manual
Table 2-1
Comparison of Displacements
16 File Reference
Example 3
Description
The plate, shown in Figure 3-1, is analyzed for uniform load only. The edges along
x = 0 and x = a are simply supported, the edge along y = b is free, and the edge along
y = 0 is fully fixed. An explicit analytical expression for the deflected surface is
given in Timoshenko and Woinowsky (1959). The deflections obtained from SAFE
are compared with the theoretical values.
The geometrical description and material properties of this problem are the same as
that of the Example 1. However, the boundary conditions differ. The detailed prob-
lem is described in the following section and shown in Figure 3-1. The gridlines
used to generate the finite element model are shown in Figure 1-2. In this example
only one load case is considered.
Data
Plate size, a b = 360 240
Plate thickness, T = 8 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.3
Load Cases:
Uniform load, q = 100 psf
Data 17
SAFE Verification Manual
Modeling Procedure
To test convergence, the problem is analyzed employing three mesh sizes, 4 4 ,
8 8 , and 12 12 , as shown in Figure 1-2. The plate is modeled using thin plate
elements available in SAFE. The two simply supported edges are modeled as line
supports with large vertical stiffnesses. The fixed edge is modeled as a line support
with large vertical and rotational stiffnesses. The self-weight of the plate is not in-
cluded in the analysis.
Comparison of Results
The numerical solution obtained from SAFE is compared with the theoretical solu-
tion which is given by Lvy (Timoshenko and Woinowsky, 1959). Comparison of
deflections shows monotonic convergence to the theoretical values with successive
mesh refinement as depicted in Table 3-1. It is to be noted that even with a coarse
mesh (4 4) the agreement is very good.
File Reference
The files for the 4 4 , 8 8 , and 12 12 meshes are S03a.FDB, S03b.FDB, and
S03c.FDB, respectively. These files are included in the SAFE package.
18 Modeling Procedure
Example 3 Rectangular Plate with Mixed Boundary
Figure 3-1
Rectangular Plate with Two Edges Simply Supported,
One Edge Fixed and One Edge Free
File Reference 19
SAFE Verification Manual
Table 3-1
Comparison of Displacements
20 File Reference
Example 4
Description
The plate, shown in Figure 4-1, is analyzed for a uniformly distributed surface load.
The edges along x = 0 and x = a are simply supported, and the other two edges are
supported on elastic beams. It is assumed that the beams resist bending in vertical
planes only and do not resist torsion. A theoretical solution to this problem is given
in Timoshenko and Woinowsky (1959). The deflections of the plate and the mo-
ments and shears of the edge beams are compared with both the theoretical solution
and the results obtained using the Direct Design Method as outlined in ACI 318-95
for = 4 . is the ratio of the bending stiffness of the beam and the bending stiff-
ness of the slab of width equal to the length of the beam and is given by the follow-
ing equation.
EI b
= , where,
aD
Eh 3
D= ,
12(1 )2
a is the length of the beam which is also equal to the one side of the slab, and
Description 21
SAFE Verification Manual
Data
Plate size, a b = 360 240
Plate thickness, T = 8 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.3
Beam Moment of Inertia, Ib = 67520 in 4
Modeling Procedure
To test convergence of results, the problem is analyzed employing three mesh sizes,
4 4 , 8 8 , and 12 12 , as shown in Figure 1-2. The slab is modeled with thin
plate elements. The simply supported edges are modeled as line supports with a
large vertical stiffness and zero rotational stiffness. Beam elements, with no tor-
sional rigidity, are defined on edges y = 0 and y = b. The flexural stiffness of edge
beams is expressed as a factor of the plate flexural stiffness.
The subdivision of the plate into column and middle strips and also the definition of
tributary loaded areas for shear calculations comply with ACI 318-95 provisions
and shown in Figure 4-2. A load factor of unity is used and the self-weight of the
plate is not included in the analysis.
Comparison of Results
As seen in Table 4-1, comparison of SAFE deflections for = 4 shows monotonic
convergence to the theoretical values with successive mesh refinement.
The variation of bending moment in edge beams along its length is shown in Table
4-2 for = 4. The theoretical solution as well as the ACI approximation by the Di-
rect Design Method is also shown. The quantity is analogous to the ACI ratio
l l (refer to Sections 13.6.4.4 and 13.6.5.1 of ACI 318-95). The correlation
1 2 1
between the numerical results from SAFE and the theoretical results is excellent.
For design purposes, the ACI approximation (Direct Design Method) compares
well with the theory. The moments were obtained at the grid points. In obtaining
SAFE moments, averaging was done at the grid points.
22 Data
Example 4 Rectangular Plate on Elastic Beams
l l = 4.3958, it is determined that the column strip supports 85% of the total
1 2 1
positive moment. The beam and slab do not carry any negative moment in the long
direction because of the simply supported boundary condition. From ACI section
13.6.5.1, for l l = 4.3958, it is determined that the beam carries 85% of the total
1 2 1
column strip moment. Since one beam supports only one-half of the column strip,
the maximum beam positive moment is 0.36125 (= 0.85 0.85 0.5) times M 0
which is equal to 975.375 kip-in. The beam moments at other locations are obtained
assuming a parabolic variation along the beam length.
Table 4-3 shows the variation of shear in edge beams for = 4. The agreement is ex-
cellent. The ACI values are calculated based on the definition of loaded tributary
areas given in Section 13.6.8.1 of ACI 318-95. The shear forces were obtained at
the middle of the grid points. In obtaining SAFE moments, no averaging was -
required for the shear forces.
File Reference
The files for the 4 4 , 8 8 , and 12 12 meshes are S04a.FDB, S04b.FDB, and
S04c.FDB, respectively. These files are included in the SAFE package.
File Reference 23
SAFE Verification Manual
a = 30'
b = 20' X
T = 8"
Material Properties :
6
Modulus of Elasticity = 3x10 psi
Poisson's Ratio = 0.3
Loading :
Figure 4-1
Rectangular Plate on Elastic Beams
24 File Reference
Example 4 Rectangular Plate on Elastic Beams
Column Strip
Edge Beam
Column Strip
Definition of Strips
Edge Beam
Figure 4-2
Definition of Slab Strips and
Tributary Areas for Shear on Edge Beams
File Reference 25
SAFE Verification Manual
Table 4-1
Comparison of Displacements
26 File Reference
Example 4 Rectangular Plate on Elastic Beams
0 0 0 0 0 0
Table 4-2
Variation of Bending Moment in an Edge Beam ( = 4)
File Reference 27
SAFE Verification Manual
Table 4-3
Variation of Shear in an Edge Beam ( = 4)
28 File Reference
Example 5
Description
The plate, shown in Figure 5-1, is analyzed for uniform load. The overall dimen-
sions of the plate are large in comparison with the column spacings a and b. Analy-
sis is limited to a single interior panel because it can be assumed that deformation is
identical in panels away from the boundaries. An analytical solution, based on the
foregoing assumption, is given in Timoshenko and Woinowsky (1959).
The numerically computed deflection, local moments, and local shears obtained
from SAFE are compared with their theoretical counterparts in Tables 5-1 through
5-3. The average design strip moments obtained from SAFE are compared with
those obtained from two ACI alternative methods, the Direct Design Method and
the Equivalent Frame Method, and the theoretical method. The comparison is
shown in Table 5-4.
Data
Plate size, a b = 360 240
Plate thickness, T = 8 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.3
Data 29
SAFE Verification Manual
Modeling Procedure
Three mesh sizes, 4 4 , 8 8 , and 12 12 , as shown in Figure 1-2, are used to test
the convergence property of the model. The model consists of a plate of uniform
thickness supported at four corners point. The effect of column support within a fi-
nite area is not modeled. Due to symmetry, the slope of the deflection surface in the
direction normal to the boundaries are zero along the edges and the shearing force
are zero at all points along the edges of the panel except at the corners. To model
this boundary condition, line supports with a large rotational stiffness about the
support line are defined on all four edges. Additional point supports are provided at
the corners. The plate is modeled with the thin plate elements in SAFE. In doing so,
the effect of shear distortion is neglected.
To approximately model the rigid corners, the slab thickness is increased for the
12 12 mesh to five times its original value in the region concerned, shown as the
40 40 areas in Figure 5-2.
To obtain design moments, the panel is divided into three strips both ways, two
column strips and one middle strip, based on the ACI 318-95 definition of design
strip widths, as shown in Figure 5-2 and in Figure 5-3. A load factor of unity is used.
The self-weight of the plate is not included in the analysis.
Comparison of Results
The numerical and the theoretical deflections are compared in Table 5-1. This table
shows monotonic convergence of the numerical solution to the theoretical values
with successive mesh refinement. The SAFE results for local moment and shear
also compare closely with the theoretical values as shown in Table 5-2 and Table
5-3, respectively. EFM is used to calculate the interior span moments as depicted in
Figure 5-2 and Figure 5-3. In Table 5-4, average strip moments obtained from
SAFE are compared with both the ACI and the theoretical values. The effect of cor-
ner rigidity is shown in Table 5-5. The agreement between the SAFE and the
theoretical solution is excellent. ACI approximations, employing either DDM or
EFM, however, deviate from the theory. It should be noted that, regardless of the
method used, the absolute sum of positive and negative moments in each direction
adds up to the total static moment in that direction.
File Reference
The files for the 4 4 , 8 8 , and 12 12 meshes are S05a.FDB, S05b.FDB, and
S05c.FDB, respectively. Also the file for studying corner rigidity is S05d.FDB.
These files are included in the SAFE package.
30 Modeling Procedure
Example 5 Infinite Flat Plate on Equidistant Columns
Floor Plan
Point Support
Corners Only
Point Support
Corners Only
A Typical Bay
Figure 5-1
Rectangular Plate on Equidistant Columns
File Reference 31
SAFE Verification Manual
Table 5-1
Comparison of Displacements
32 File Reference
Example 5 Infinite Flat Plate on Equidistant Columns
Location Moments
(kip-in/in)
Mx My
Table 5-2
Comparison of Local Moments
File Reference 33
SAFE Verification Manual
Shear Force
(10 3 kip/in)
Location Vx Vy
Table 5-3
Comparison of Local Shears
34 File Reference
Example 5 Infinite Flat Plate on Equidistant Columns
Corner Stiffening
Column Strip
X
Middle Strip
Column Strip
Figure 5-2
Definition of X-Strips (Moment values are obtained by EFM)
File Reference 35
SAFE Verification Manual
Corner Stiffening
Y Column Strip
Middle Strip
Column Strip
Figure 5-3
Definition of Y-Strips (Moment values are obtained by EFM)
36 File Reference
Example 5 Infinite Flat Plate on Equidistant Columns
Table 5-4
Comparison of Average Strip Moments
File Reference 37
SAFE Verification Manual
Not computed
Table 5-5
Comparison of Average Strip Moments : Effect of Corner Rigidity
38 File Reference
Example 6
Description
An infinite plate resting on elastic subgrade and carrying equidistant and equal
loads, P, is shown in Figure 6-1. Each load is assumed to be distributed uniformly
over the area u v of a rectangle. A theoretical double series solution to this exam-
ple is given in Timoshenko and Woinowsky (1959).
The numerically computed deflections, local moments, and local shears obtained
from SAFE are compared to the theoretical values as shown in Table 6-1 and Table
6-2.
Data
Plate size, a b = 360 240
Plate thickness, T = 15 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.2
Modulus of subgrade
reaction, k = 1 ksi/in
Loading: Point Load, P = 400 kips
(assumed to be uniformly distributed over an area u v)
Data 39
SAFE Verification Manual
Modeling Procedure
Analysis is confined to a single interior panel. Three mesh sizes, 4 4 , 8 8 , and
12 12 are used to model the panel as shown in Figure 1-2. The slab is modeled
with thin plate elements and the elastic support is modeled as a surface support with
a spring constant of k, the modulus of subgrade reaction. The edges are modeled as
line supports with a large rotational stiffness about the support line. Point loads P 4
are defined at the panel corners. In the theoretical formulation (Timoshenko and
Woinowsky 1959), each column load P is assumed to be distributed over an area
u v of a rectangle, as shown in Figure 6-1. To apply the theoretical formulation to
this problem, concentrated corner loads are modeled as a uniformly distributed load
acting over a very small rectangular area where u and v are very small.
Comparison of Results
Excellent agreement has been found between the numerical and theoretical deflec-
tion for k = 1 ksi/in as shown in Table 6-1. As the modulus k is changed, the distri-
bution of pressure between the plate and the subgrade changes accordingly. The
particular case, as k approaches 0, corresponds to a uniformly distributed subgrade
reaction, i.e., to the case of a reversed flat slab uniformly loaded with q = P ab. In
fact the problem changes to that of Example 5, with the direction of vertical axis re-
versed. In Example 5, for a uniform load of 100 psf (P = 60 kips), the maximum
relative deflection is calculated as 0.280 . Applying the formulation used here with
k = 1 10 yields a deflection value of 0.279 . SAFE local moments using the
6
12 12 mesh have been compared with the theoretical results in Table 6-2. The
results agree well.
File Reference
The files for the 4 4 , 8 8 , and 12 12 meshes are S06a.FDB, S06b.FDB, and
S06c.FDB, respectively. These files are included in the SAFE package.
40 Modeling Procedure
Example 6 Infinite Flat Plate on Elastic Subgrade
FLOOR PLAN
A TYPICAL BAY
Material Properties :
Modulus of Elasticity = 3000 ksi
Poissons Ratio = 0.2
Subgrade Modulus =1 ksi/in
Loading :
Typical Column Load = 400 kips
Figure 6-1
Rectangular Plate On Elastic Subgrade
File Reference 41
SAFE Verification Manual
Table 6-1
Comparison of Displacements
42 File Reference
Example 6 Infinite Flat Plate on Elastic Subgrade
Moments
(kip-in/in)
Location Mx My
Table 6-2
Comparison of Local Moments
File Reference 43
.
Example 7
Description
A skew plate under uniform load, as shown in Figure 7-1, is analyzed for two differ-
ent support configurations. In the first case, all the edges are assumed to be simply
supported. In the second case, the edges y = 0 and y = b are released, i.e., the plate is
assumed to be supported on its oblique edges only. A theoretical solution to this
problem is given in Timoshenko and Woinowsky (1959). In both cases, the maxi-
mum deflection and the maximum moment are compared with the corresponding
theoretical values.
Data
Plate size, a b = 480 240
Plate thickness, T = 8 inches
Modulus of elasticity, E = 3000 ksi
Poissons ratio, = 0.2
Modeling Procedure
A 8 24 base mesh is used to model the plate as shown in Figure 7-1. A large verti-
cal stiffness is defined for supports, and support-lines are added on all four edges
Modeling Procedure 45
SAFE Verification Manual
for the first case and along the skewed edges only for the second case. A load factor
of unity is used. The self-weight of the plate is not included in the analysis.
Comparison of Results
The comparison of SAFE and the theoretical results is excellent, as shown in Table
7-1. Under the simply supported boundary condition, maximum deflection occurs
at the plate center and the maximum principal moment acts nearly in the direction
of the short span. Under the simply supported condition on the oblique edges and
free boundary conditions on the other two edges, maximum deflection occurs at the
free edge as expected.
File Reference
The files for two different boundary conditions are S07a.FDB and S07b.FDB.
These files are included in the SAFE package.
46 Comparison of Results
Example 7 Skew Plate with Mixed Boundary
Figure 7-1
Skew Plate
File Reference 47
SAFE Verification Manual
Maximum displace-
0.156 0.162
ment (inches)
Simply supported
on all edges
Maximum bending
3.66 3.59
moment (kip-in)
Maximum displace-
ment at the free edges 1.51 1.50
(inches)
Simply supported
on oblique edges
Maximum bending
moment at the free 12.16 11.84
edges (kip-in)
Displacement at the
1.21 1.22
center (inches)
Simply supported
on oblique edges Maximum bending
moment at the center 11.78 11.64
(kip-in)
Table 7-1
Comparison of Deflections and Bending Moments
48 File Reference
Example 8
Description
The flat slab system arranged three-by-four is shown in Figure 8-1. The slab con-
sists of twelve 7.5 inch thick 18 22 panels. Edge beams on two sides extend 16
inches below the slab soffit. Details are shown in Figure 8-2. There are three sizes
of columns and in some locations, column capitals. Floor to floor heights below and
above the slab are 16 feet and 14 feet respectively. A full description of this prob-
lem is given in Example 1 of ACI 340.R-97 (ACI Committee 340, 1997). The total
factored moments in an interior E-W design frame obtained from SAFE are com-
pared with the corresponding results obtained by the Direct Design Method, the
Modified Stiffness Method, and the Equivalent Frame Method.
Data
Materials:
Data 49
SAFE Verification Manual
Loading:
Modeling Procedure
The computational model uses a 10 10 mesh of elements per panel, as shown in
Figure 8-3. The mesh contains gridlines at column centerlines, column faces, and
the edges of column capitals. The grid-lines extend to the slab edges. The regular
slab thickness is 7.5 . A slab thickness of 21.5 is used to approximately model a
typical capital. The slab is modeled with thin plate elements. The columns are mod-
eled as point supports with vertical and rotational stiffnesses. Stiffness coefficients
used in the calculation of support flexural stiffness are all reproduced from ACI
Committee 340 (1997). Beams are defined on two slab edges as shown in Figure
4
8-1. The value of torsional constant (10767 in ) used for the beams is also repro-
duced from ACI Committee 340 (1997).
The model is analyzed for uniform factored load of 0.365 ksf (wu = 1.4wd + 1.7 wl )
in total including self-weight. To obtain factored moments in an E-W interior de-
sign frame, the slab is divided into strips in the X-direction (E-W direction) as
shown in Figure 8-4. An interior design frame consists of one column strip and two
halves of adjacent middle strips.
Comparison of Results
The SAFE results for the total factored moments in an interior E-W design frame
are compared in Figure 8-5 with the results obtained by the Direct Design Method
(DDM), the Modified Stiffness Method (MSM), and the Equivalent Frame Method
(EFM). Only uniform loading with load factors of 1.4 and 1.7 has been considered.
The DDM, MSM, and EFM results are all reproduced from Example 1 of ACI
Committee 340 (1997), the Alternative Example 1 of ACI Committee 340 (1991),
and from Example 3 of ACI Committee 340 (1991), respectively. Moments re-
ported are calculated at the face of column capitals. Overall, they compare well. A
noticeable discrepancy is observed in the negative column moment in the west side
of the exterior bay (the edge beam side). In contrast to the EFM, the DDM appears
to underestimate this moment. The SAFE result falls in between the two extreme
values. The basic cause of this discrepancy is the way in which each method ac-
counts for the combined flexural stiffness of columns framing into the joint. The
DDM uses a stiffness coefficient k c of 4 in the calculation of column and slab flex-
50 Modeling Procedure
Example 8 ACI Handbook Flat Slab Example 1
ural stiffnesses. The EFM, on the other hand, uses higher value of k c to allow for
the added stiffness of the capital and the slab-column joint. The use of MSM affects
mainly the exterior bay moments which is not the case when the DDM is employed.
In SAFE, member contributions to joint stiffness are dealt with more systematically
than any of the above mentioned approaches. Hence, the possibility of overdesign-
ing or underdesigning a section is greatly reduced.
The factored strip moments are compared in Table 8-1. There is a discrepancy in
the end bays, particularly on the edge beam (west) side, where the SAFE and EFM
results for exterior negative column strip moment show the greatest difference.
This is expected because EFM simplifies a 3D structure to a 2D structure, thereby
neglecting the transverse interaction between adjacent strips. Except for this local-
ized difference the comparison is good.
File Reference
The file for this example is S08.FDB which is included in the SAFE package.
File Reference 51
SAFE Verification Manual
Edge Beam
Design Frame
Edge Beam
Figure 8-1
Flat Slab From ACI Handbook
52 File Reference
Example 8 ACI Handbook Flat Slab Example 1
Figure 8-2
Sections and details of ACI Handbook Flat Slab Example
File Reference 53
SAFE Verification Manual
Figure 8-3
SAFE Mesh (10 10 per panel)
54 File Reference
Example 8 ACI Handbook Flat Slab Example 1
Figure 8-4
Definition of E-W Design Frames and Strips
File Reference 55
SAFE Verification Manual
Figure 8-5
Comparison of Total Factored Moments (E-W Design Frame)
56 File Reference
Example 8 ACI Handbook Flat Slab Example 1
DDM 6 62 54 43 37 43 48 57 0
MSM 10 55 52 43 37 43 46 48 0
Middle
Strip
EFM 10 55 53 48 29 48 54 41 0
SAFE 29 73 46 43 50 39 42 64 2
Table 8-1
Comparison of Total Factored Strip Moments (kip-ft)
(Interior E-W Design Frame)
File Reference 57
.
Example 9
Description
The two-way slab system arranged three-by-three is shown in Figure 9-1. The slab
consists of nine 6.5 inch thick 20 24 panels. Beams extend 12 inches below the
slab soffit. Details are shown in Figure 9-2. 16 16 columns are used throughout
the system. Floor to floor height is 15 ft. A full description of this problem is given
in Example 2 of ACI 340.R-91 (ACI Committee 340, 1991). The total factored mo-
ments in an interior design frame obtained from SAFE are compared with the Di-
rect Design Method, the Modified Stiffness Method, and the Equivalent Frame
Method.
Data
Concrete strength, f c = 3 ksi
Yield strength of steel, fy = 40 ksi
Concrete unit weight, wc = 150 psf
Modulus of elasticity, Ec = 3120 ksi
Poissons ratio, = 0.2
Data 59
SAFE Verification Manual
Modeling Procedure
The computational model uses a 10 10 mesh of elements per panel, as shown in
Figure 9-3. The mesh contains grid lines at both column centerlines and column
faces. The grid lines are extended to the slab edges. The slab is modeled with thin
plate elements. The columns are modeled as point supports with vertical and rota-
tional stiffnesses. A stiffness coefficient of 4EI L is used in the calculation of sup-
4 4
port flexural stiffness. Torsional constants of 4790 in and 5478 in are defined for
the edge and interior beams respectively, in accordance with Section 13.7.5 of ACI
318-89 and Section 13.0 of ACI 318-95 code. The model is analyzed for uniform
factored total load of 0.347 ksf ( wu = 1.4wd + 1.7 wl ) including self-weight. To ob-
tain factored moments in an interior design frame, the slab is divided into strips in
X-direction (E-W direction) as shown in Figure 9-4. An interior design frame con-
sists of one column strip and two halves of adjacent middle strips.
Comparison of Results
The SAFE results for the total factored moments in an interior E-W design frame
are compared with the results obtained by the Direct Design Method (DDM), the
Modified Stiffness Method (MSM), and the Equivalent Frame Method (EFM) as
shown in Figure 9-5. The results are for uniform loading with load factors. The re-
sults are reproduced from ACI Committee 340 (1991). Moments reported are cal-
culated at the column face. For all practical purposes they compare well. At the end
bays, the MSM appears to overestimate the exterior column negative moments with
the consequent reduction in the mid-span moments.
The distribution of total factored moments to the beam, column strip, and middle
strip is shown in Table 9-1. The middle strip moments compare well. The total col-
umn strip moments also compare well. The distribution of the column strip mo-
ments between the slab and the beam has a larger scatter.
File Reference
The file for this example is S09.FDB which is included in the SAFE package.
60 Modeling Procedure
Example 9 ACI Handbook Two-way Slab Example 2
Figure 9-1
ACI Handbook Two-way Slab Example
File Reference 61
SAFE Verification Manual
Figure 9-2
Details of Two-way Slab Example from ACI Handbook
62 File Reference
Example 9 ACI Handbook Two-way Slab Example 2
Figure 9-3
SAFE Mesh (10 10 per panel)
File Reference 63
SAFE Verification Manual
Figure 9-4
Definition of E-W Design Frames and Strips
64 File Reference
Example 9 ACI Handbook Two-way Slab Example 2
120
191 191
Figure 9-5
Comparison of Total Factored Moments (kip-ft)
in an Interior E-W Design Frame
File Reference 65
SAFE Verification Manual
DDM 9 23 28 25 14 25
Slab MSM 13 21 28 25 14 25
Column
Strip EFM 12 21 30 27 11 27
SAFE 24 27 62 58 14 58
DDM 3 69 84 76 41 76
Slab MSM 5 63 84 76 41 76
Middle
Strip EFM 4 63 89 82 34 82
SAFE 8 65 72 71 46 71
Table 9-1
Comparison of Total Factored Moments (kip-ft)
66 File Reference
E x a m p l e 10
Description
This example models the flat plate structure tested by the Portland Cement Associa-
tion (Guralnick and LaFraugh, 1963). The structure consists of nine 5.25 thick
15 15 panels arranged 3 3 as shown in Figure 10-1. Deep and shallow beams
are used on the exterior edges. The structure is symmetric about the diagonal line
through columns A1, B2, C3, and D4, except the columns themselves are not sym-
metric about this line. The corner columns are 12 12 and the interior columns
are 18 18 . Columns along the edges are 12 18 with the longer dimension
parallel to the plate edge. A typical section of the plate and details of edge beams are
given in Figure 10-2. The total moments in an interior frame obtained numerically
from SAFE are compared with the test results and the numerical values obtained by
the Equivalent Frame Method (EFM).
Data
Concrete strength, f c = 4.1 ksi
Yield strength of steel, fy = 40 ksi
Concrete unit weight, wc = 150 pcf
Modulus of elasticity, Ec = 3670 ksi
Poissons ratio, = 0.2
Data 67
SAFE Verification Manual
Modeling Procedure
A finite element model, shown in Figure 10-3, with 6 6 mesh per panel is em-
ployed in the analysis. The slab is modeled using the thin plate elements in SAFE.
The columns are modeled as point supports with vertical and rotational stiffnesses.
The reduced-height columns in the test structure are fixed at the base. Hence, rota-
tional stiffnesses of point supports are calculated using a stiffness coefficient of 4
and an effective height of 39.75 inches (K c = 4EI / lc ). The calculation of torsional
stiffness of edge beams is based on the requirements of Section 13.7.5 of ACI 318-
95. The model extends to the centerlines of the edge columns only. The portion of
slab occupying the column area is modeled as rigid by increasing its thickness to
five times the nominal slab thickness. A total uniformly distributed design load of
156 psf (not factored) is applied to all the panels.
To obtain design moment coefficients, the plate is divided into column and middle
strips. An interior design frame consists of one column strip and half of each adja-
cent middle strip. Normalized values of design moments are used in the compari-
son.
Comparison of Results
The SAFE results for the total non-factored moments in an interior frame are com-
pared with test results and the Equivalent Frame Method (EFM). The test and EFM
results are all obtained from Corley and Jirsa (1970). The moments are compared in
Table 10-1. The negative design moments reported are at the faces of the columns.
Overall, the agreement between the SAFE and EFM results is good. The experi-
mental negative moments at exterior sections, however, are comparatively lower.
This may be partially the result of a general reduction of stiffness due to cracking in
the beam and column connection at the exterior column which is not accounted for
in an elastic analysis. It is interesting to note that even with an approximate repre-
sentation of the column flexural stiffness, the comparison of negative exterior mo-
ments between EFM and SAFE is excellent.
File Reference
The file for this example is S10.FDB which is included in the SAFE package.
68 Modeling Procedure
Example 10 PCA Flat Plate Test
Design Frame
Shallow Beam Side
Design Frame
Deep Beam Side
Figure 10-1
PCA Flat Plate Example
File Reference 69
SAFE Verification Manual
Figure 10-2
Section and Details of PCA Flat Plate Example
70 File Reference
Example 10 PCA Flat Plate Test
Figure 10-3
SAFE Mesh (6 6 per panel)
File Reference 71
SAFE Verification Manual
PCA Test 0.037 0.047 0.068 0.068 0.031 0.073 0.073 0.042 0.031
EFM 0.044 0.048 0.067 0.062 0.038 0.062 0.068 0.049 0.043
SAFE
0.044 0.050 0.068 0.062 0.041 0.062 0.067 0.051 0.042
(Shallow Beam Side)
SAFE
0.043 0.050 0.067 0.062 0.041 0.061 0.067 0.050 0.042
(Deep Beam Side)
* Wl = 526.5 kip-ft
1
Table 10-1
Comparison of Measured and Computed Moments
72 File Reference
E x a m p l e 11
Data
Material:
Concrete strength, f c = 2.5 ksi
Yield strength of steel, fy = 36.7 ksi
Modulus of elasticity, Ec = 2400 ksi
Poissons ratio, = 0.2
Loading:
Total uniform load w = 140 psf
Data 73
SAFE Verification Manual
Modeling Procedure
The computational model uses a 6 6 mesh of elements per panel as shown in
Figure 11-3. The mesh contains gridlines at column centerlines as well as column
faces. The slab is modeled with thin plate elements and the columns are modeled as
point supports with vertical and rotational stiffnesses. The reduced-height columns
in the test structure are pinned at the base. Hence, an approximate value of 3
(K c = 3EI lc ) is used to calculate flexural stiffness of the supports taking the col-
umn height as 9.5 . The slab is thickened over the column sections to account for ri-
gidity of the slab-column joint. Shallow and deep beams are defined on the edges
4 4
with torsional constants of 16 in and 14.9 in , respectively, as described in Section
13.0 of ACI 318-95. The model is analyzed for uniform total load of 140 psf.
To obtain maximum factored moments in an interior design frame, the plate is di-
vided into columns and middle strips. An interior design frame consists of one col-
umn strip and half of each adjacent middle strip.
Comparison of Results
The SAFE results for uniform load moments for an interior frame are compared
with the experimental and EFM results in Table 11-1. The experimental and EFM
results are all obtained from Corley and Jirsa (1970). The negative design moments
reported are at the faces of the columns. From a practical standpoint, even with a
coarse mesh the agreement between the SAFE and EFM results is good. In general
the experimentally obtained moments at exterior sections are low, implying a loss
of stiffness in the beam-column joint area.
In comparing absolute moments at a section, the sum of positive and average nega-
tive moments in the bay should add up to the total static moment. The SAFE and
EFM results comply with this requirement within an acceptable margin of accu-
racy. The experimental results are expected to show greater discrepancy because of
the difficulty in taking accurate strain measurements.
File Reference
The file for this example is S11.FDB which is included in the SAFE package.
74 Modeling Procedure
Example 11 University of Illinois Flat Plate Test F1
Design Frame
Shallow Beam Side
Design Frame
Deep Beam Side
Figure 11-1
University of Illinois Flat Plate Test F1
File Reference 75
SAFE Verification Manual
Figure 11-2
Sections and Details of University of Illinois Flat Plate Test F1
76 File Reference
Example 11 University of Illinois Flat Plate Test F1
Figure 11-3
SAFE Mesh (6 6 per panel)
File Reference 77
SAFE Verification Manual
+ + +
M M M M M M M M M
TEST F1 0.027 0.049 0.065 0.064 0.040 0.058 0.058 0.047 0.034
EFM 0.047 0.044 0.072 0.066 0.034 0.067 0.073 0.044 0.046
SAFE
0.044 0.050 0.068 0.062 0.041 0.062 0.067 0.051 0.042
(Shallow Beam Side)
SAFE
0.044 0.050 0.067 0.062 0.041 0.061 0.066 0.050 0.042
(Deep Beam Side)
* Wl = 17.5 kip-ft
1
Table 11-1
Comparison of Measured and Computed Moments
78 File Reference
E x a m p l e 12
Description
This example models, F2 and F3, the flat slab structures tested by the University of
Illinois by Hatcher, Sozen, and Siess (1969) and Jirsa, Sozen, and Siess (1966) re-
spectively. A typical structure used in tests F2 and F3 is shown in Figure 12-1. The
fundamental difference between these two test structures is in the type of reinforce-
ment used. In test F2, the slab is reinforced with medium grade reinforcement
whereas in test F3 welded wire fabrics are used. The structure consists of nine
5 5 panels arranged three-by-three. Two adjacent edges are supported by deep
beams, 2 in. wide by 6 in. deep, and the other two edges by shallow beams, 4.5 in.
wide by 2.5 in. deep, producing a single diagonal line of symmetry through col-
umns A1, B2, C3, and D4. A typical section and details of columns, drop panels,
and column capitals are shown in Figure 12-2. For both structures, the numerical re-
sults obtained for an interior frame by SAFE are compared with the experimental
results and the EFM results due to uniformly distributed load.
Data
Concrete strength:
Data 79
SAFE Verification Manual
Modulus of elasticity:
Poissons ratio:
= 0.2
Loading:
Modeling Procedure
The computational model uses a 8 8 mesh of elements per panel as shown in Fig-
ure 12-3. The mesh contains gridlines at the column centerlines as well as the edges
of drop panels and interior column capitals. The mesh extends to the centerlines of
the edge columns only. In the absence of edge beams, it is recommended to include
an extra line of elements beyond the edge column centerlines up to the slab edge.
The slab thickness is increased to 2.5 inches over the drop panels. A thickness of 4.5
inches is used to approximately model the interior capitals. Short deep beams are
used to model the edge column capitals. In this model, the slab is modeled with thin
plate elements and the columns are modeled as point supports with vertical and ro-
tational stiffnesses. A stiffness coefficient of 4.91 (K c = 4.91 EI c / lc ) is used in the
calculation of the support flexural stiffness based on a column height of 21.375
inches, measured from the mid-depth of the slab to the support center. Due to the
presence of capitals, columns are treated as non-prismatic. Shallow and deep beams
4 4
are defined on the edges with torsional constants of 15.4 in and 17.8 in respec-
tively as described in Section 13.0 of ACI 318-95.
Both the test problems are modeled in SAFE with concrete modulus of elasticity of
2100 ksi. This affects the slab, beam, and column stiffnesses. Since the distribution
of moment depends on the relative stiffnesses, the test problems are not modeled
twice, one for E c = 2100 ksi and the other for E c = 3700 ksi.
The model is analyzed for uniform load. To obtain maximum factored moments in
an interior design frame, the slab is divided into two interior and two exterior design
80 Modeling Procedure
Example 12 University of Illinois Flat Slab Tests F2 and F3
Comparison of Results
The SAFE results for moments in an interior frame, are compared with the experi-
mental and EFM results for both structures F2 and F3 in Table 12-1. The experi-
mental and EFM results are all obtained from Corley and Jirsa (1970). Moments
are compared at the edge of column capitals. Table 12-1 shows that the SAFE and
the EFM results are in excellent agreement. In general, the measured positive mo-
ments appear to be lower than the SAFE and EFM values.
File Reference
The file for this example is S12.FDB which is included in the SAFE package.
File Reference 81
SAFE Verification Manual
Design Frame
Shallow Beam Side
Design Frame
Deep Beam Side
Figure 12-1
University of Illinois Flat Slab Tests F2 and F3
82 File Reference
Example 12 University of Illinois Flat Slab Tests F2 and F3
Figure 12-2
Sections and Details of Flat Slabs F2 and F3
File Reference 83
SAFE Verification Manual
Figure 12-3
SAFE Mesh (8 8 per mesh)
84 File Reference
Example 12 University of Illinois Flat Slab Tests F2 and F3
TEST F2 0.025 0.042 0.068 0.062 0.029 0.061 0.065 0.038 0.025
TEST F3 0.029 0.038 0.057 0.055 0.023 0.058 0.060 0.034 0.024
EFM 0.021 0.044 0.057 0.050 0.026 0.049 0.057 0.044 0.021
SAFE
0.028 0.044 0.062 0.056 0.029 0.055 0.061 0.045 0.026
(Shallow Beam Side)
SAFE
(Deep Beam Side) 0.028 0.043 0.062 0.055 0.029 0.055 0.061 0.044 0.026
* Wl = 35.0 kip-ft
1
Table 12-1
Comparison of Measured and Computed Moments
File Reference 85
.
E x a m p l e 13
Description
This example models the slab structure tested by the University of Illinois by Gam-
ble, Sozen, and Siess (1969). The structure is a two-way slab, 1.5 in. thick, in which
each panel is supported along all four edges by beams as shown in Figure 13-1. The
structure consists of nine 5 5 panels arranged three-by-three. The edge beams
extend 2.75 below the soffit of the slab and the interior beams have an overall
depth of 5 . The corner columns are 4 4 and the interior columns are 6 6 .
Edge columns are 4 6 with the longer dimension parallel to the slab edge. A
typical section of the slab and details are shown in Figure 13-2. The moments in an
interior design frame due to uniform loads obtained from SAFE are compared with
the corresponding experimental results and the numerical values obtained from the
EFM.
Data
Concrete strength, f c = 3 ksi
Yield strength of reinforcements, fy = 42 ksi
Modulus of elasticity, Ec = 3000 ksi
Poissons ratio, = 0.2
Data 87
SAFE Verification Manual
Modeling Procedure
The computational model uses a 6 6 mesh of elements per panel as shown in Fig-
ure 13-3. Gridlines are defined at column faces as well as the column centerlines.
The mesh extends to the edge column centerlines only. The slab is modeled using
the thin plate elements available in SAFE. The columns are modeled as supports
with both vertical and rotational stiffnesses. A stiffness coefficient of 8.0 is used in
the calculation of support flexural stiffnesses based on a column height of 15.875 ,
measured from the mid-depth of the slab to the support center. The column is as-
sumed to be infinitely rigid over the full depth of the beams framing into it. The
value of 8.0 is 75% of the figure obtained from Table 6.2 of ACI Committee 340
(1997) to approximately account for the pinned end condition at the column base.
The slab is also thickened over the column sections to approximately model rigidity
4 4
of the slab-column joint. Torsional constants of 33.7 in and 24.2 in are defined for
the interior and edge beams respectively, in accordance with Section 13.0 of ACI
318-95.
To obtain maximum factored moments in an interior design frame, the slab is di-
vided into two interior and two exterior design frames spanning in the X-direction
(E-W direction). Because of double symmetry, comparison is confined to X-strips
only. An interior design frame consists of one column strip and half of each adja-
cent middle strip.
Comparison of Results
The SAFE results for moments in an interior frame are compared with experimental
and EFM results in Table 13-1. The test and EFM results are all obtained from Cor-
ley and Jirsa (1970). The negative design moments reported are at the face of col-
umns. The comparison is excellent. The minor discrepancy is attributed to the loss
of stiffness due to the development of cracks and the difficulty in measuring strains
accurately at desired locations.
File Reference
The file for this example is S13.FDB which is included in the SAFE package.
88 Modeling Procedure
Example 13 University of Illinois Two-way Slab Test T1
Figure 13-1
University of Illinois Two-way Slab Example T1
File Reference 89
SAFE Verification Manual
Figure 13-2
Sections and Details of Slab T1
90 File Reference
Example 13 University of Illinois Two-way Slab Test T1
Figure 13-3
SAFE Mesh of Slab T1 (6 6 per panel)
File Reference 91
SAFE Verification Manual
* Wl = 18.75 kip-ft
1
Table 13-1
Comparison of Measured and Computed Moments
92 File Reference
E x a m p l e 14
Description
This example models the slab structure tested by the University of Illinois by Van-
derbilt, Sozen, and Siess (1969). The structure is a two-way slab arranged three-
by-three panels in which each panel is supported along all four edges by beams as
shown in Figure 14-1. The structure consists of nine 1.5 thick 5 5 panels. The
edge beams and the interior beams extend 1.5 below the soffit of the slab. The cor-
ner columns are 4 4 and the interior columns are 6 6 . Edge columns are
4 6 with the longer dimension parallel to the slab edge. A typical section of the
slab and details are shown in Figure 14-2. The moments in an interior design frame
obtained numerically from SAFE are compared with the experimental results and
the EFM results.
Data
Concrete strength, f c = 3 ksi
Yield strength of reinforcement, f y = 47.6 ksi
Modulus of elasticity, Ec = 3000 ksi
Poissons ratio, = 0.2
Data 93
SAFE Verification Manual
Modeling Procedure
The computational model uses a 6 6 mesh of elements per panel as shown in
Figure 14-3. Gridlines are defined at column faces as well as the column centerli-
nes. The mesh extends to the edge column centerlines only. The slab is modeled
with thin plate elements and the columns are modeled as supports with both vertical
and rotational stiffnesses. A stiffness coefficient of 6.33 is used in the calculation of
support flexural stiffnesses based on a column height of 13.125 in., measured from
the mid-depth of the slab to the support center. The column stiffness is assumed to
be infinitely rigid over the full depth of the beams framing into it. The value of 6.33
is 75% of the figure obtained from Table A7 of Portland Cement Association
(1990) to approximately account for the pinned end condition at the column base.
The slab is also thickened over the column section to allow for rigidity of the slab-
4 4
column joint. Torsional constants of 11.2 in and 10.6 in are defined for the inte-
rior and edge beams respectively, in accordance with Section 13.0 of ACI 318-95.
To obtain maximum factored moments in an interior design frame, the slab is di-
vided into two interior and two exterior design frames spanning in the X-direction
(E-W direction). An interior design frame consists of one column strip and half of
each adjacent middle strip.
Comparison of Results
The SAFE results for moments in an interior frame are compared with the experi-
mental and EFM results in Table 14-1. The experimental and EFM results are all
obtained from Corley and Jirsa (1970). The negative design moments reported are
at the face of columns. The comparison is excellent except for the negative exterior
moments where the experimental results are lower than both the SAFE and the
EFM results. The discrepancy is attributed not only to the loss of stiffness due to the
development of cracks, but also to the difficulty in taking accurate strain measure-
ments at desired locations.
File Reference
The file for this example is S14.FDB which is included in the SAFE package.
94 Modeling Procedure
Example 14 University of Illinois Two-way Slab Test T2
Figure 14-1
University of Illinois Two-Way Floor Slab T2
File Reference 95
SAFE Verification Manual
Figure 14-2
Sections and Details of Floor Slab T2
96 File Reference
Example 14 University of Illinois Two-way Slab Test T2
Figure 14-3
SAFE Mesh of Slab T2 (6 6 per panel)
File Reference 97
SAFE Verification Manual
* Wl = 17.375 kip-ft
1
Table 14-1
Comparison of Measured and Computed Moments
98 File Reference
E x a m p l e 15
A one-way simple-span slab supported by walls on both long edges is modeled us-
ing SAFE. The slab consists of a 6 inches thick 12 feet wide panel and is shown in
Figure 15-1. To ensure one-way action Poissons ratio is taken to be zero. The slab
moment on a strip of unit width is computed analytically. The total factored strip
moments are compared with the SAFE results. These moments are identical. After
the analysis was done, design was performed using the ACI 318-95 code by SAFE
and also by hand computation. The design reinforcements computed by the two
methods are also compared in Table 15-1.
Description 99
SAFE Verification Manual
Data
Thickness, T, h = 6 in
Depth of tensile reinf., dc = 1 in
Effective depth, d = 5 in
Depth of comp. reinf., d = 1 in
Clear span, ln , l 1
= 144 in
Length, l2
= 720 in
Modeling Procedure
The computational model uses a finite element mesh, automatically generated by
SAFE. The maximum element size was specified to be 48 inches. To obtain fac-
tored moments and flexural reinforcement in a design strip, one one-foot wide strip
is defined in X-direction on the slab as shown in Figure 15-1. The slab is modeled
with thin plate elements. The walls are modeled as line supports without rotational
stiffnesses and with very large vertical stiffness (1 10 Kip/in.).
20
One dead load (DL80) and three live load (LL000, LL100, LL800) cases with uni-
formly distributed surface loads of magnitudes 80, 0, 100, and 800 psf, respec-
tively, are defined in the model. Three load combinations (COMB000, COMB100,
and COMB800) are defined with the ACI 318-95 load combination factors, 1.4 for
dead loads and 1.7 for live loads. The model is analyzed for these load cases and
load combinations.
Calculation of Reinforcement
The following quantities are computed for all the load combinations:
= 0.90
b = 12 in
100 Data
Example 15 Design Verification of Slab
f 4000
= 0.85 0.05 c = 0.85
1000
1
87 000
cb = d = 2.959 in
87 000 + f y
Mu = 1
wl = 0 psf
w = 9.333 lb/in
M u = 24.192 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 0.1335 in < amax
0.85 f c b
f y d
2
wl = 100 psf
w = 23.5 lb/in
M u = 60.912 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 0.3436 in < amax
0.85 f c b
fy d
2
As = 0.2336 sq-in
wl = 800 psf
w = 122.67 lb/in
M u = 317.952 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 2.2283 in > amax
0.85 f c b
a
M uc = C d max = 281.018 kip-in
2
Therefore the moment resisted by compression steel and tensile steel is:
M us = M u M uc = 36.934 kip-in
The stress in the compression steel is given by:
a d
f s = 0.003 E s max 1
= 47.8 ksi < f y
amax 1
The required tensile steel for balancing the compression in concrete is:
M uc
As = = 1.2828 sq-in
1
amax
f y (d )
2
The tensile steel for balancing the compression in steel is given by:
M us
As = = 0.1710 sq-in
f y ( d d )
2
As = As + As = 1.4538 sq-in
1 2
Comparison of Results
The SAFE total factored moments in the design strip are compared with the
moments obtained by the analytical method in Table 15-1. They match exactly for
this problem. The design reinforcements are also compared in Table 15-1.
File Reference
The file for this example is S15.FDB which is included in the SAFE package.
12' 0''
Free edge
60' 0'' X
Free edge
Figure 15-1
One-way Slab
Figure 15-2
Strip Reinforcement for Medium Load Levels
Reinforcement Area
Moment (sq-in)
Load Level Method
(kip-in)
As+ As
* A+ s, min
= 0.1296 sq-in
Table 15-1
Comparison of Design Moments and Reinforcements
The stress-block remains within the flange and the computed tensile reinforce-
ment falls below the minimum permitted in ACI,
The stress-block remains within the flange, the computed tensile reinforcement
exceeds the minimum permitted in ACI, and remains within the balanced con-
dition permitted by ACI,
The stress-block extends below the flange but remains within the balanced con-
dition permitted by ACI,
The stress-block extends below the flange and exceeds the permitted balanced
condition, requiring compression reinforcement.
Description 107
SAFE Verification Manual
Data
Clear span, l = 240 in
Overall depth, h = 18 in
Flange Thickness, ds = 4 in
Width of web, bw = 12 in
Width of flange, bf = 24 in
Depth of tensile reinf., dc = 3 in
Effective depth, d = 15 in
Depth of comp. reinf., d = 3 in
Modeling Procedure
The computational model uses a finite element mesh of frame elements, automati-
cally generated by SAFE. The maximum element size was specified to be 6 inches.
The beam is supported by columns without rotational stiffnesses and with very
large vertical stiffness (1 10 Kip/in).
20
One dead load (DL03) and four live load (LL00, LL10, LL30, LL40) cases with
only symmetric third-point loads of magnitudes 3, 0, 10, 30, and 40 kips, respec-
tively, are defined in the model. Three load combinations (COMB00, COMB10,
COMB30, and COMB40) are defined with the ACI 318-95 load combination fac-
tors of 1.4 for dead loads and 1.7 for live loads. The model is analyzed for all of
these load cases and load combinations.
Calculation of Reinforcement
The following quantities are computed for all the load combinations (Please refer to
the SAFE Users Manual for more details):
= 0.90
108 Data
Example 16 Flexural Design Verification of Beam
3 f
c 200
As max bw d and bw d = 0.60 sq-in
f y fy
f 4000
= 0.85 0.05 c = 0.85
1000
1
87 000
cb = d = 8.8776 in
87 000 + f y
Pu = 1.4 Pd + 1.7 Pl
Pu L
Mu =
3
Pl = 0 kip
Pu = 4.2 kips
M u = 336 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 0.3082 in (a < amax and a < d s )
0.85 f c b f
f y d
2
Pl = 10 kips
Pu = 21.2 kips
M u = 1696 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 1.6279 in (a < amax and a < d s )
0.85 f c b f
f y d
2
As = 2.2140 sq-in
Pl = 30 kips
Pu = 55.2 kips
M u = 4416 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 4.7658 in (a > d s )
0.85 f c b f
Calculation for As is done in two parts. The first part is for balancing the compres-
sive force from the flange,C f , and the second part is for balancing the compressive
force from the web, C w . C f is given by:
Therefore, the area of tensile steel reinforcement to balance flange compression is:
M uf
As = = 2.7200 sq-in
f y ( d d s 2)
1
M uw = M u M uf = 2506.56 kip-in
The web is a rectangular section of dimensions bw and d, for which the design depth
of the compression block is recalculated as
2 M uw
a = d d 2
= 5.5938 in (a amax )
0.85 f c bw
1 1
The area of tensile steel reinforcement to balance the web compression is then
given by:
M uw
As = = 3.8037 sq-in
2
a
f y d 1
As = As + As = 6.5237 sq-in
1 2
Pl = 40 kips
Pu = 72.2 kips
M u = 5776 kip-in
The depth of the compression block is given by:
2 Mu
a= d d
2
= 6.7719 in (a > d s )
0.85 f c b f
Calculation for As is done in two parts. The first part is for balancing the compres-
sive force from the flange,C f , and the second part is for balancing the compressive
force from the web, C w . C f is given by:
Therefore, the area of tensile steel reinforcement to balance flange compression is:
M uf
As = = 2.7200 sq-in
f y ( d d s 2)
1
M uw = M u M uf = 3866.56 kip-in
The web is a rectangular section of dimensions bw and d, for which the design depth
of the compression block is recalculated as:
2 M uw
a = d d 2
= 11.2048 (a > amax )
0.85 f c bw
1 1
Therefore the moment resisted by concrete web and tensile steel is:
a
M uc = C d max = 2529.1619 kip-in
2
M us = M uw M uc = 1337.3981 kip-in
As = As + As + As = 8.6323 sq-in
1 2 3
Comparison of Results
The SAFE total factored moments in the beam for different load combinations are
compared in with the moments obtained by the analytical method. They match ex-
actly for this problem. The design reinforcements are also compared in Table 16-1.
They also match exactly.
File Reference
The file for this example is S16.FDB which is included in the SAFE package.
P P
PL/3
24''
3'' 4''
18''
3''
12''
Beam Section
Figure 16-1
The Model Beam for Flexural Design
Reinforcement Area
Moment (sq-in)
Load Level Method
(kip-in)
As+ As
Table 16-1
Comparison of Moments and Flexural Reinforcements
The average shear stress in the beam falls below half of the concrete capacity,
requiring no shear reinforcement,
The average shear stress in the beam falls above half of the concrete capacity
but not exceeding the concrete capacity by 50 psi, requiring minimum shear re-
inforcement according to the ACI 318-95 code,
The average shear stress in the beam falls below the maximum shear stress al-
lowed by ACI 318-95, requiring design shear reinforcement,
The average shear stress in the beam exceeds the maximum shear force allowed
by ACI 318-95 representing a failure condition.
Description 117
SAFE Verification Manual
Data
Clear span, l = 240 in
Overall depth, h = 18 in
Flange Thickness, hf = 4 in
Width of web, bw = 12 in
Width of flange, bf = 24 in
Modeling Procedure
The computational model uses a finite element mesh of frame elements, automati-
cally generated by SAFE. The maximum element size was specified to be 6 inches.
The beam is supported by columns without rotational stiffnesses and with very
large vertical stiffness (1 10 Kip/in).
20
One dead load (DL03) and four live load (LL00, LL10, LL30, LL60) cases with
only symmetric third-point loads of magnitudes 3, 0, 10, 30, and 60 kips, respec-
tively, are defined in the model. Three load combinations (COMB00, COMB10,
COMB30, and COMB60) are defined with the ACI 318-95 load combination fac-
tors of 1.4 for dead loads and 1.7 for live loads. The model is analyzed for all of
these load cases and load combinations.
Calculation of Reinforcement
The following quantities are computed for all the load combinations (Please refer to
the SAFE Users Manual for more details):
= 0.85
118 Data
Example 17 Shear Design Verification of Beam
Vc = 2 f c bw d = 19.353 kips
Vs = 8 f c bw d = 77.41 kips
If Vu (Vc 2) ,
Av
=0,
s
Pu = 1.4 Pd + 1.7 Pl
Vu = Pu
Pl = 0 kip
Pu = 4.2 kips
Pl = 10 kips
Pu = 21.2 kips
Pl = 30 kips
Pu = 55.2 kips
Pl = 60 kips
Pu = 106.2 kips
Comparison of Results
The SAFE total factored shear-forces in the beam for different load combinations
are compared in Table 17-1 with the shear-forces obtained by the analytical
method. They match exactly for this problem. The design shear reinforcements are
also compared in Table 17-1. They also match exactly.
File Reference
The file for this example is S17.FDB which is included in the SAFE package.
P P
PL/3
24''
3'' 4''
18''
3''
12''
Beam Section
Figure 17-1
The Model Beam for Shear Design
Av
Reinforcement Area,
s
Shear Force
Load Level (sq-in/ft)
(kip)
SAFE Calculated
Table 17-1
Comparison of Shear Reinforcements
125
SAFE Verification Manual
126
Notes